Download fibrillation complicating acute myocardial

CLINICAL RESEARCH
European HeartJournat (2014) 35, 116-122
A r rh rlth m i o I
doi:1 0.1 093/eurheartj/eht45 3
ele ctr o phy si o I o gy
lncidence of sudden cardiac death after ventricular
fibrillation complicating acute myocardial
infarction: a S-year causê-of-death analysis of the
FAST-MI 2005 registryr
Wulfran Bougouinl,z,3, Eloi Marijonl'2,3, Etienne Puymiratl'2'3, Pascat Defayea,
David S. Celermaiers, Jean-Yves Le Heuzèy1'2, Serge Boveda6, Salem KacetT,
Phitippe Mabo8,e, Claude Barnayr0, Antoine Da Costall, fean-Claude Deharol2,
Jean-Claude Dauberts,e, f ean Ferrièresl3, Tabassome Simonl4'ls,
and Nicolas Danchinl'2'3*, on behalf of FAST-Ml Registry lnvestigators
Paris, France
Received 12 June
2013; revised
I
August 201
3: occ9ted 28 September 201 3; online publish<heodof-print
l9
November 201 3
V
""j:j: :
'1.:.
'::;,
,we
rge, according
* Corresponding author. TeL +33
1
56 09 2471,Fu: +33 1 56 09 25 72, EmaiL [email protected]
iPresented in part (mid-term interim FAST-Ml analysis at 3 yeare) at the Annual Congresses ofthe American Heart Association (Orlando, FL, USA, 12-16 November 2011) and
the European Society of Cardiology (Paris, France, 27-31 August 201 1), and presented in full at the Annual Congress of the American Heart Associatlon (Daltas, TX, USA,
1
6-20 November
201 3).
Published on behatf of the European Society of Cardiology. Alt righs reserued. @ The Author 201 3. For pemissions please emaiL [email protected]
1t7
lncidence of SCD after ventricular fibriltation
lntroduction
The prognosis fottowing acute myocardiat infarction (Ml) has improved
considerabty in recent decades, principatty due to the devetopment of
percutaneous coronary intervention and improved medical therapy'
inctuding thrombotytic agents.l'2 However' despite these major
advances, ventricutar fibrittation (VF) or rapid ventricular tachycardia
(Vl) stitt occur during the acute phase of Ml in 2-8% of cæes'3-6
Seveml studies have shown that Patients who devetop VF during
the acute phase of Ml have a higher risk of death in the short-term'4's'7
ln contrast, data on mid-to-long-term suryivalare [imited and controversial.s-13 ln the long-term, the mode of death is poorty characterized, particulartythe riskof sudden cardiac death (SCD). Despite this
area of significant uncertainty, the European and North American
guidetines in 2008 did not recommend early implantation of imptantabte cardioverter-defibritlators (lCD) in such patients, even though
experts agreed on the particutarty low tevel of evidence in the fietd
(tevet of evidence C;.14 The main reasons behind those guidetines
were the reversible nature of the arrhythmic trigger (acute myocardiatischaemia) and the presumed efficacyof currenttreatments (coronary revascularization) in preventing recurrences ofVF. However,
recent studies have questioned the dogma that such treatment of reversible causes of malignant arrhythmia can be sufficient to Prevent
later fatal event.ls-20 ln this setting, data addressing long-term fotlowup
as
wettas specific causes of death in survivors ofVF are necessaryto
whether such patients are at higher long-term risk of SCD.21
assess
ln the present analysis,
we rePort the S-year cause-of-death ana-
with M l, according to whether or not they devetoped
their index hospitat stay.
> 48 h
exclusion criteriawere: (i) refusal to consent; (2) Ml admitted
within
occurring
Mls
as
after symptoms; (3) iatrogenic Mls, defined
angioptasty'
coronary
surgery,
(bypass
48 h of a therapeutic Procedure
Acute Ml with
ing to the current guidetines at the time of study initiation'
(STEMI) was defined as acute M I with ST elevation > 1 mm in
Si elevation
qualifying
at least two contiguous leads in any location in the index or
any
without
Ml
acute
as
Ml
ECG; and non-ST-segment-etevation
died
who
Patients
ECG'
quatifying
or
the
index
in
elevation
ST-segment
very earty after admission and for whom cardiac biomarkers were not
measured were included if they had compatibte signs or symptoms associated with tyPical ST changes.
lnthe presentanalysis, we studied patientswith Ml accordingto whether
or not they devetoped VF duringthe acute phase (defined by VF occurring
after diagnosis of Ml is made). Ventricutarfibrittation was defined as irregutar waves of inconsistent shaPe without distinct QRS comptexes or
T-waves. Ventricutar fibritlation was ctassified as earty VF if it occurred in
the first 48 h of admission, and late VF if it occurred after the first 48 h
but before hospitat discharge. Patients who devetoped stable VT, even
with haemodynamic compromise requiring urgent cardioversion, were
not considered in the VF group. ln an additional analysis, we considered
patients who developed stabte VT, defined as mpid (>120 b.p.m) ventrjcular rate with no features for associated atrial tachycardia' lasting
more than 30 s, with or without haemodynamic compromise requiring
urgent cardioversion. We assessed the influence of VF on the S-year
fottow-up,
as
well
as
the combined influence of VT associated to
VF.
tysis of patients
VF during
Methods
Study setting and population
The methodotogy of the FAST-MI 2005 Registry has been described in
detail elsewhere.2'22'" B.iefly, the primary objective was to evaluate
acute Ml management in 'real life' Pl?ctice, and to assess medium and
long-term prognosis of patients admitted to intensive care units (lCUs)
after acute Ml. This registry comes from a prospective mutticentre
study (223 centres) including 3670 Ml Patients' Patients were recruited
consecutively from ICU departments over a period of 1 month (October
2005), with an additional 1 -month extension for diabetic patients. Pafticipation in the study was offered to alt French institutions, university teaching hospitats, general and regional hospitals, and private ctinics with lCUs
to receive acute coronary syndrome emergencies' ln each
centre, a physician was in charge ofthe registry and provided a full list
of all patients admitted to his/her unit. The study was conducted in compliance with Good Clinicat Practice, French Law and the French data protection law. The protocol was reviewed by the Committee for the
Protection of Human Subiects in Biomedical Research (CCPPRB) of
Saint-Antoine University Hospital. Datafile of the FAST-Ml 2005 registry
was declared to and authorized bythe French data Protection committee
authorized
(Commission Nationate lnformatique et Liberté' CNIL).
The inclusion criteria were: (i) man or woman aged over 18 years;
(ii) patient admifted within 48 h of symptom onset in an ICU for an
acute Ml characterized by the elevation ofcardiac biomarkers associated
with at least one of the fotlowing elements: symPtoms compatibte with
myocardial ischaemia, onset of pathotogical Q waves, ST-T changes compatibte with myocardia[ ischaemia (ST segment etevation or depression,
T-wave inversion); and (iii) having agreed to take Part in the study. The
Annuat follow-up and mode of death
adjudication Process
A modified Hinkte-Thater system was used to classif, deaths into SCD,
non-SCD, and non-cardiovascular death.2a Sudden cardiac death was
assumed in patients who: (i) died suddenty and unexpectedty within t h
ofcardiac symptoms in the absence ofprogressive cardiac deterioration:
(ii) died unexpectedly in bed during sleep; or (iii) died unexpectedly
within 24 h after last being seen ative. Other cardiovascular deaths
included Ml, heart faiture, acute aortic syndrome, and putmonary embotism. Termina[ arrhythmias associated with heart failure deaths were clas-
sified as non-sudden cardiovascutar deaths. Deaths that could be
attributed to non-cardiovascular reasons were ctassified as noncardiovascutar death (cancer, infectious disease, rena[ faiture, respiratory
faiture). lf avaitabte datawere insufficientto make a reasonabte identification of the cause of death, it was considered as unctassifiable' When no
data were available, cause of death was adjudicated æ unknown.
Local investigators notified deaths and their causes during the index
hospital stay, using a predefined detaited form. A specific working
group ensured annual fotlow-uP of Patients, using specific etectronic
Case Report Forms through contacts with the attendÎng physicians'
and/or the patients themselves for additional information. Follow-up
focused particularly on vital status, specific mode, and cause of death,
as we[[ as significant medical events or interyentions during follow-up
(especiatty regarding modification of pharmacotogical treatment, hosPitatization, or ICD implantation). lf missing' vital status was obtained
from the registries ofthe patients' birthptaces, and from French Nationat
lnstitute
of
Health and Medical Research-INSERM CépiDc Unit
(Kremtin-Bicêtre, France). Overalt, 3-year fotlow-up was completed in
97.8%of patients, and S-yearfoltow-up in94.5% of Patients.
A systematic review of att death notifications was performed' with
central adjudication of alt events, by two independent cardiologists
1t8
W. Bougouin et ol.
btinded to VF status. ln cases ofdivergent opinion on the mode ofdeath, a
third expertwas asked to arbitrate. Primary endpoints ofthe presentanatysis were overall mortatity and cause-specific mortality among hospital
survivors of acute Ml according to whether or not they devetoped VF
during the acute Ml admission.
chronic kidney disease, anaemia), chronic cardiovascutar medications
before the acute event, Ml characteristics (STEMI vs. non-STEMI),
12 lead-ECG at admission (teft bundte btock branch, atrial fib illation),
revascutarization procedures at admission (medical, percutaneous cor-
onary intervention, coronary artery bypass), anti-thrombotic during
the first 48 h (tow-motecular weight heparin, gtycoprotein llb/llla antago-
Statistical analysis
nists), extent ofcoronary artery disease, left-ventricutarejection fraction,
discharge medications, heart rate, and systolic blood pressure at discharge. A backward stepwise selection was apptied to obtain a finat
modelthat included covariates with P < 0.05. Att tests were two-taited,
and values of P < 0.05 were considered significant. Statistical anatysis
was performed using SPSS 18.0 software (SPSS, lnc., Chicago, lL, USA).
This repoft was prepared in comptiance with the STROBE checktist for
observational studies.2s Normality assumption was checked for continuous variables, and continuous variables were presented as mean
standard deviation, and discrete variables were presented as percentages.
Comparisons between groups (patients with or without VF) were made
with 12 or Fisher's exact tests for discrete variables, and with unpaired
t-tests or one-way ANOVA for continuous variabtes. ln-hospital survivat
was described as percentages and comparisons anatysed using 12 test.
Factors associated with occurrence ofVFwere identified using multivariate
*
Results
lncidence of VF and associated factors
backw-ard stepwise logistic regression anatysis.
Among the 3670 enrolted patients, 1'16 devetoped VF during the
acute phase of Ml, giving an incidence ræe of 3.2% (95% C\2.6-3.8),
among Ml patients ative at hospital admission. This group inctuded
92 early VF (79.3"Â) and 24 tate VF (20.7%). The mean time from Ml
Overall as well as specific mortality rates were calcutated. Survivat
curyes over the S-year foltow-up period were estimated using the
Kaptan-Meier estimation. Cox proportionat hazards regression anatysis
was used to assess variabtes independently associated with overalt mortatity. The cumulative hazard functions for each covariable were computed to assess proportionatity. Variabtes with a value of p < 0.10 in
to occurrence of VF was 1.8 days (95% Cl
diagnosis
ïoble
1.4
-2.2).
univariate analyses were included in the final multivariate model, inctud-
shows baseline characteristics of patients, especiatty medicat
history and previous treatments, accordingto VF occurrence during the
ing demographic characteristics, comorbidities (cardiovascutar risk
factors including body mass index, heart failure, stroke history cancer,
acute phase of Ml. Patients with VF were younger, presented with
tess frequent history of systemic arterial hypertension (47.4 vs. 60%,
Baseline characteristics
VF (n
Age, year (SD)
Male sex, n (%)
BMl, kg/m'z (sD)
=
7
116)
No VF (n = 3554)
P-value
63.3 (14.4)
67.4 (13.e)
0.002
88 (7s.e)
2427 (68.3)
0.084
27.3 (s.0)
27.2 (4.8)
Medical history
:t:
Diabetes meltitus, n (%)
34 (2e.3)
1282 (36.1)
0.135
Hypertension, n (%)
ss (47.4)
2132 (60.0)
0.007
Hypertipidaemia, n (%)
4s (38.8)
1729 (48.6)
0.037
Current smoker, n (%)
4s (38.8)
1020 (28,7)
0,018
COPD, n (%)
s (4.3)
174 (4.e)
0.77
Chronic kidney disease, n (%)
4 (3.4)
20s (s.8)
0.286
Prior Ml, n (%)
le
Prior CABG, n (%)
(16.4)
647 (18,2)
0.616
4 (3.4)
206 (s.8)
0.284
3 (1 1.2)
sos (14.2)
0.361
188 (s.3)
0.051
Prior PCl, n (%)
1
Prior stroke, n (%)
11 (e.s)
Prior transient ischaemic attack n (%)
4 (3.4)
131(3.7)
0.89
Prior peripheral vascular disease, n (%)
e (7.8)
3se (10.1)
0.403
Prior congestive heart faiture, n (%)
s (4.3)
20e (s.e)
0.474
0.266
Previous treatment
Beta-btockers, n (%)
24 (20.7)
8e7 (2s.2)
Aspirin, n (%)
23 (1e.8)
e11 (2s.6)
0.158
e (7.8)
470 (13.2)
0.085
Clopidogret, n (%)
Statin, n (%)
26 (22.4)
1
00s (28.3)
0.167
ACE-inhibitor or ARB, n (%)
28 (24.'t)
1
26s (3s.6)
0.01
124 (3.s)
0.601
Amiodarone, n (%)
3 (2.6)
1
119
lncidence of SCD after ventricutar fibriltation
:
:
according to the occurrence of VF are shown in Figure 1 ' ln-hospitat
mortatity was higher among patients with VF than in patients
0.007) and hypertipidaemia (38.8 vs.48,6%, P 0'037)' and were
more often current smokers (38.8 vs. 28.7%' P:0.018)' Previous
P
without VF (25.0 vs. 5'0%, P < 0.001). After adjustment by age' sex'
type of Ml, cardiovascutar risk factors, history of stroke, history of
treatment with angiotensin converting enzyme (ACE) inhibitors or
angiotensin recePtor btocker (ARB) was tess frequent in the VF
group (24.1 vs. 35.6%, P 0.011). Conversety, previous treatment by
chronic kidney disease, heart rate, btood pressure, atriat fibritlation
at admission, medications used before entry, and use of coronary
angiography during the hospitat stay, in-hospital mortality remained
higher among patients with VF than Patients without VF (adiusted
OR 7.38, 95"/" Cl 4.27-12.75, P < 0'001). When considering VF
timing, in-hospital mortatity was significantty higher in the tate VF
:
beta-blockers did not differ according to VF occurrence (20.7 vs.
25.2%, P :0.266). Other treatments, such as antiptatetet agents or
antiarrhythmics, did not differ between groups. Overatt' 518 pæients
(14.1%) presented a history of percutaneous coronary intervention
at baseline. Characteristics of Ml and detaits on management are summarized in Toble 2. ln Patients undergoing percutaneous coronary intervention, Sg% in the VF group and79% in the non-VF group had TlMl 3
groupthan in the eartyVFgroup (33'3 vs.22.8%,P < 0.001) (Frgure 1)'
Ioble 3 summarizes causes of in-hospitat death. Causes of deaths
significantty differed between VF and non-VF patients (P < 0'001)'
Whereas cardiogenic shock represented 37.9% of in-hospital
flow at the end of the procedure (P: 0.14). Among Patients who
underwent percutaneous coronary intervention, 48 (41.4%) and
1270 (35.7%) had singte-vesset 'disease (P:0.46)' in the VF and
non-VF groups, respectively. Anterior Ml (vs. non-anterior) was signifi-
causes of death in the non-VF group (simitarthan SCD)' cardiogenic
shock was the predominant cause of death among non-VF patients
(55.1%). When considering only in-hospital SCD, most of the
deaths in the VF group (81.8%) were arrhythmia-related with unsuccessful resuscitation, whereas 61.5% of such deaths in the non-VF
group were retated to non-arrhythmic SCD, principatly cardiac
rupture and tamponade.
cantlyassociated to the occurrence ofVF in univariate anatysis (37.9% in
YF vs.20.1% in non-VF, P
<
0.001).
ln multivariate anatysis, independent factors associated with early
occurrence of VF were STEMI (OR 2.4,95% C11.5-3.6, P < 0.001),
age
(60
years (OR1.7,95% Cl
1.1
-2.5,
P
:
0.01), history of stroke
(OR2.4,95%C|1.2-4.7,P: 0.01), and atrialfibrittation
on first ECG
(OR 2.5, 95% Cl 1.4-4.4, P : 0.003).
Long-term outcome of patients alive
at hospital discharge
ln-Hospital mortality and cause.of-death
Treatments at hospital discharge are described in Ïoble
analysis
were not statistically different between VF patients and non-VF
Overatt, 207 deaths occurred during the hospitat stay, giving
a
rates
VF (n
Characteristics of Ml
Treatments
patients regarding beta-btockers, aspirin, ctopidogrel, statin, and atdosterone btockers. ln contrast, Patients from the VF group were
mortal-
ity rate of 5.6% (95"/o Cl 4.8-6.4). ln-hospital mortatity
4.
= l16)
No VF (n = 3554)
Clinical characteristics
<3
<0.001
e3 (80.2)
2097 (se.s)
4Q6.2)
83s (23.6)
126.2 (31.8)
140.4 (28.s)
83.7 (24.8)
80.1 (20.2)
84 (72.4)
1788 (s0.3)
4 (3.4)
1ss (4.4)
Anterior Ml, n (%)
44 (37.e)
713 (20.1)
Atrialfibrillation on first ECG, n (%)
1s (12.e)
2se (7.3)
0.023
0.015
Symptom onset to calt
Kiltip ctass
>
h, n (%)
1, n (%)
Systotic btood pressure, mmHg (SD)
Heart rate, b.p.m. (SD)
STEMI, n (%)
Left bundle branch btock
n (%)
0.002
<0.00'1
0.067
<0.001
0.635
<0,001
ParaclinicaI
Admission glycaemia, g/L, mean (SD)
1.8 (0.e)
1.6 (0.8)
Admission LDL, g/1, mean (SD)
1.1 (0.s)
1.2 (0.4)
White blood cetl count,
1Oe/L, mean (SD)
GMCE score, mean (SD)
0.378
<0.001
12.s (4.4)
10.3 (3.8)
110.9 (32.6)
11s.e (32.4)
0.117
0.072
ln-hospital course
83 (71.6)
2253 (63.4)
CABG, n (%)
2 (1.7)
144 (4.1)
ICD implantation during hospitat phase, n (%)
4 (3.4)
8 (0.2)
<0.001
e.3 (8.1)
<0.001
Pct,n(%)
Length of hospitat stay, days (SD)
12.3 (1s.2)
0.328
W. Bougouin et ol.
120
Boo
50
s
.à
f
.à
40
E
Ë
6
Ë
E30
6
40
o
E30
o
E
.=
=o-
8cn
o-
t-'
Êto
s
s
Patients without
VF
Patients with VF
VF
(n =
NoVF
87)
(n = 3376)
(84.7)
(e3.1)
72 (82.8)
73 (84.e)
6e (81.2)
Beta-btockers, n (%)
Aspirin, n (%)
Ctopidogrel, n (%)
Statins, n (%)
ACE-inhibitor or
72
2s82 (77.e)
0.134
81
308e (e2.2)
0.745
2707 (80.e)
0.667
2788 (83.e)
0.802
2318 (70.4)
0.031
ARB, n (%)
Aldosterone btocker,
6
(6.e)
164 (4.e)
0.32
(8)
(13.8)
183 (s.4)
0.333
2s1(7.4)
0.027
n (%)
Orat anticoagutant, n (%)
Amiodarone, n (%)
more often discharged with amiodarone (1 3.8 vs.7.4%, P : 0.027)
and ACE-inhibitor/ARB (81.2 vs. 70.4%, P:0.031) than patients
7
12
(95% Cl 73.1-89.5%) in the VF group compared with 74.2% (95o/"
C\72.6-75.8%) in the non-VF group (Frgure 2). ln muhivariate ana-
without VF. Left-ventricutar ejection fraction was lower at discharge
lysis, after adjustment
among VF patients than non-VF patients (46.6 vs. 52.0%, P < 0.001
).
The rate of ICD imptantation during in-hospitatcoursewas low (0.3%
VF duringthe acute phase of Ml was not associated
in
the overall poputation), but higher in the VF group (3.4
vs. 0.2%, p
<
After 3 years of fottow-up (avaitabte in 97.8% of patients),
pharmacotogicat therapies no longer differed between the two
0.001).
groups, including antiarrhythmic medications, particu tarty amiodarone (P: 0.22). An additional 35 patients received an ICD during
for other prognostic factors, occurrence of
with any increase
(HRfor mortatity 0.78, 95%Cl0.38-L58, p :
0.21), in both STEMI and non-STEMI patients. When considering
timing of VF, neither earty VF (OR 0.5, 95% Cl 0.2-1.1, P: 0.09)
in mortatity at5 years
:
nor tate VF (OR 1.9,95%C10.6-5.9, P 0.30) was associated significantty with tong-term mortatity. When we considerthe additionalT5
cases of
VTto the VF group, in muttivariate
anatysis, after adjustment
the first 3 years of foltow-up, atl of these in the non-VF group,
for other prognostic factors, occurrence of sustained ventricu(ar
giving an overall ICD imptantation rate of 1 .2%, not significantty differ-
arrhythmias (VT and VF considered together) during the acute
phase of Ml was not associated with significant increase in mortality
at 5 years (HR 0.95, 95% Cl 0.64-1.42, P:0.81). Comparativety
ent between VF (3.4%) and non-VF (1.2%) groups.
Among 3463 suruivors at hospitat discharge, 1024 died during a
mean follow-up of 52 + 2 months, giving an overatt survival rate at
5 years of 74.4% (95% Cl 72.8-76.0%). Survivat rates were 81.3%
to
patients without any sustained ventricutar arrhythmias, neither
patients with W onty (HR 1.79,95% Cl 0.79-2.11, P: 0.31) nor
l2l
lncidence of 5CD after ventricular fibrittation
Patlents wlth VF
F - - af r
- r r r r - -
\-
i,
:::
Patlentc wlthout VF
Ë
rl:
::
aE
60C,6
j
Adlu*ted HR 0.78,
ccr/'
9596
Cl0.38-1.5t
:,::
:
.
:)
0%
ii
Ntnntlc.atd8k
Pâùênts wlth
I
Dkcharge
VF
87
Psti€nlswithoÉVF 3il7t
1
year
2
3
Yearc
2Wùg
3 fqq.
80'
78
v4
s053
2873
270s
4yêani
tt
2d?É
'':,
patients with VF onty (HR 0.65,95%
increased 5-year mortality.
Cl 0.32-1.33,
p:
unclassifiabte causes of death was not statisticat(y different between
the two groups (30.4 vs.29.8%). Among overatl ascertained deaths,
most were from cardiovascutar causes (65%), inctuding SCD (29%
of deaths from cardiovascutar cause) and non-SCD (72o/o of deaths
from cardiovascular cause). No significant difference was found in
the distribution of causes of death in the VF and non-VF groups
(P : 0.71). The proportion of deaths r.elated to SCD was 13.1% in
the VF group compared wirh 12.9% in the non-VF group (p:
Discussion
(23t87)
From this mutticentre and unsetected cohort of more than 3500 Ml
out in the modern era of reper-fusion therapy, we
report, tothe best ofour knowledge, the first tong-term cause-of-death
accordingtothe devetopment ofVF duringthe acute stage. We
the classical predictors of VF and the high nte of
in-hospitat mortatity among patients who developed VF, especialty
from SCD. However, our resutts suggest that among patients disanalysis
have confirmed
charged ative from hospitat after Ml, the survivat rate at 5 years did
not differ significantty between VF and non-VF patients. Furthermore,
incidence rate of SCD was low, and remar*abty similar in both groups.
Coronary heart disease remains the most common cause undercases.26
NoVF
(t001/3376)
071
Sudden cardiac death, n (%)
Non-sudden cardiac
(13.1) 12e (12.e)
s (21.7) 331 (33.1)
3
I
death, n (%)
Non-cardiovascular
8
(34.8)
242 (24.2)
7
(30.4)
zee (2e.8)
death, n (%)
Unknown or unclassifiable,
n (/")
event and dectines significantly thereafter, reaching a steady state
approximatety one year.28'2e Consideri ng these epidemiotogicat
findings, severat studies have evatuated the potential benefit of ICD in
aft er
patients carried
of
t >
,.:'.j
VF
0.71), with respective annual incidence rates of 0.68"Â (95% Cl
0.18-1.7 4) and Q.7 6% (95% Cl 0.64-0.90).
75"/o
55
1787
0.24) had
Causes ofdeaths in the two groups (VF and non-VF) are reported
in ïoble 5. Overalt, causes of death were determined in 71g patients
during the S-year fottow-up (70.1%). The freguency of unknown or
lying SCD in the western wortd, responsibte
6ysaæ
for
approximatety
The teading cause of VF is myocardial ischaemia,2?
and it has been demonstrated that the absotute risk offatat arrhyth_
mic event after Ml is greatest within the week immediately after the
primary prevention shortly after Ml, and found no evidence for
benefit.3o-32 ln contrast, whereas patients devetoping VF at the
acute phase of Ml might seem at highest risk of recurrence of VF,
no RCTs have evaluatedthe potential interestoflCD in such patients.
ln the setting of secondarT prevention, the low-risk profite of patients
who develop VF or haemodynamicatly unstabte VT in the acute setting
of ischaemia (<48 h) on one hand, as well as the high-risk profite of
patients who develop such arrhythmias with no recent (<40 days)
Ml and no recent coronary revascutarization (< 3 months), has been
demonstnted. ln contrast, there is an important grey zone regarding
patients in between these time periods. This has been recentty under_
lined in the 201 3 ACCF/HRS/AHA/ASE/HFSA/SCA|/SCCT/SCMR
Appropriate Use Criteria for lmptantable Cardiover-ter-Defibrittators
and Cardiac Resynchronization Therapy ReporL33
W. Bougouin et ol.
122
we identified severat previously described factors associated with the occurrence ofVF, such as STEMl,s youngerag",t'3aAF at
initiatpresentation.3s Ourfinding of a higher riskof in-hospital mortatln this study,
ity among patiens with VF in the setting of acute Ml is also consistent
1'34-36
4..ot6ingty, current guidetines advowith the literature.s'7-eJ
cate ctose monitoring (continuous ECG) for at teast 24 h after Ml, and
for up to 72 h in high-risk patients (especiatty Patients with severe
arrhythmias).37 Ourresutts supportthe importance of ctose monitoring during hospitalstay. Accordingto ourfindings, increased early mortatity in VF patients cannot be explained by less frequent adequate
revascularization, as ludged by post-procedureTlMl flow' lnterestingty'
we observed that SCD resutted from non-arrhythmia-related causes in
an important proportion of cases in Patients without VF in agreement
with the findings of Pouleur et o1.38 Our results also emphasize the
more unfavourabte in-hospitat outcomes of tate VF when compared
with early VF, consistent with previous studies.4'10'36 Taken together,
our resutts are very consistent with the existing literature regarding
of these arrhythmias seem different' when they occur in the setting
of
inctude only patients with VF in
acute
specificities, possibty prognostic'
pathophysiologicat
respect
orderto
of this arrhythmia.
M1.11
Thus, we decided
to
We acknowtedge some timitations. First, our data are observationat'
However, this study inctudes patients from different centres' avoiding
selection bias inherent in randomized controtted trials. Secondly, initial
treatment modatities coutd vary from one centre to another, and this
study reflecs reat daity clinical practice. Thirdty, cause ofdeath coutd
onty be determined in 70% of deaths. Considering the tong foltow-up
(5 years, with very few Patients lost to foltow-up), the report of causes
in
of deaths and the similarproportion of undetermined cause of death
VF and non-VF grouPs, our data can be regarded as robust'
ln conclusion, in this long-term fottow-up of a large cohort
of lYl
patients treated according to contemPorary medicat therapy, occurrence of VF in the initiat phase was associated with higher in-hospitat
mortatity. ln contrast, long-term outcome (especiatty incidence of
wetl as short-term mortatity.
We repor! to the best of our knowledge, the fim long-term
cause-of-death anatysis according to the devetopment of VF during
SCD) was not different between VF and non-VF Patients' This reinforces the crucial importance of ctose monitoring duringthe hospitalization of such patients, but a[so emphasizes the timited utitity of earty
the acute stage. To the best of our knowtedge, onty three previous
studies have reported tong-term fottow-up after VF compticating
Ml.s'13'3s Two of those studies included Patients from the prereperfusion era, before widespread use of thrombotytic and PCI treatments, etements that have undoubtedly changed the prognosis of
ICD imptantation in such Patients.
predictors of VF occurrence
as
They reported an increased short-term moftatity but suggested no prognostic imPact uP to 3 years. However, the onty
cause-of-death anatysis avaitabte to date was recentty reported by
Ml.8'34'3e
Piccini eto1.,1121 describingspecific mortalities at 30 days of 228 sublects.
Despite the paucity ofdata regarding delayed cause-specific death,
it is usuatty considered that patients with VF at the acute phase of Ml
do not have increased tong-term risk of SCD. However, ctinicat, epidemiotogical, and genetic studies recentty add substantial evidence
that patients are not equat concerning the risk of devetoping VF
during ischaemia,ls'16'18'1e and that Patients who have developed
VF during the acute phase of Ml coutd be at higher long-term risk
of further matignant arrhythmia and SCD. At this time, occurrence
of VF resulting from reversible trigger (especiatty, ischaemia at the
acute phase of Ml) does not currently rePresent an indication for
ICD implantation,ao with a verT low tevel of evidence. ln this
context, our resutts provide substantiat data, first because of the
S-year fotlow-up (tongest avaitabte to date in such a setting). Moreover, because comparabte overatt mortatity rates coutd conceat significant differences in modes of death (especiatty arrhythmic death)
during fottow-up, our cause-of-death anatysis is of particutar interest'
ln this study, we reportthat patients with VF duringthe acute phase of
Ml present with very simitar incidence rates of later SCD' despite
comparabte treatments during fottow-up, inctuding ICD imptantation
(in onty a minority of cases, argund 1%). This adds substantiat evidence to the tegitimacy of current guidelines in discouraging ICD imptantation in the setting ofVF occurring duringthe acute phase of Ml.
Untit now, most studies that evaluated short- or mid-term prognosis of Ml patients according to the occurrence of early ventricutar
arrhythmias have merged sustained W and YF.4'6'7'12 However, VF
is most often triggered by acute myocardial ischaemia,al whereas
W (atthough triggered by acute myocardiat ischaemia) usually
invotves a structural substrate.a2'a3 Furthermore, prognostic vatues
Funding
N.D. has received research grants from Astra-Zeneca' Daiichi-Sankyo'
Eti-Litty, GSK, Merck, Novartis, ffizer, Sanofi-aventis, Serrrier, and The
Medicines Company.
S.B. has received fees as a consuttant for Medtronic
and Boston Scientific; N.D. has received fees as a speaker or consuttant
Conflict of interesû
for AstraZeneca, Bayer, Bristol-Myers Squibb, Boehringer-lngelheim,
Daiichi-SanÇo, Eti-Litty, GtaxoSmithKtine, MSD-Schering' Novartis,
Novo-Nordisk, ffizer, Roche, Sanofi-Aventis, Servier, and The Medicines
Company. No other conflict of interest of other co-authors.
References
K, Wright FL, Rayner M, Goldacre MJ Determinants of the decline in mortality from acute myocardial infarction in Engtand between 2002 and 201 0: linked national database stu dy. BMJ 2012;344:d8059.
Puymimt E, Simon T, Steg PG, Schiete F, Guéret P, Blanchard D, Khatife K, Goldstein P,
Cattan S, Vaur L, Cambou J-P, FerrièresJ, Danchin N. fusociation of changes in clinical charucteristics and management with improvement in survival among Patients
with ST-elevation myocardial infarction . JAIÂA 2012l,308:998- 1 006'
1. Smolina
2.
PJ, Otteruanger JB de Boer M-J' Dambrink J-HE'
Gossetink ATM, van 't Hof AWJ, HoorntjeJCA, SurTapranata H' Zijlstm F Ventricular fibrillation in acute myocardiat infarction before and during primarT PCI'
3. Henriques JPS, Gheeraert
J Cordiol 2005 :1 O9:262 -266.
Mehta RH,StarrAZ, LoPes RD, HochmanJS'Widimsky P' Pieper KS,ArmstrongPW,
Gnnger CB. lncidence of and outcomes associated with ventricutar lachycardia or
I nt
4.
fibrillation in Patients undergoing Primary Percutaneous coronary intervention'
- 1 7 89.
5. At-Khatib SM, GrangerCB, HuangY, Lee KL, CaliffRM, Simoons ML' Armstrong PW'
Van de Werf F, White HD, Simes RJ, Moliterno DJ, Topol EJ, Harrington RA' Sustained ventricutar arrhythmias among Patients with acute coronatT syndromes
with no ST-segment etevation: incidence, Predictors' and outcomes' Grculotion
J A h4A 2OA9 30 1 :1 7 7 9
6.
2002;l 06:309-31 2.
Mehta RH, Harlai KJ, Grines l- Stone GW, BouraJ, Cox D, O'Neitl W, Grines CL
sustained ventriculartachycardia or fibrittation in the cardiac catheterization taboratory among patients receiving Primary Percutaneous coronary intervention: inci-
dence, predictors, and outcomes. J Am Coll Cordiol 2004:43:17 65 - 177 2'
7. PicciniJP,BergerJS,BrownDL,Earlysustainedventriculararrhythmiascomplicating
acute myocardiat infarction. Am I Med 7008:12l:7 97 -804.
8. JensenGV,Torp-PedersenC,HitdebrandtBKoberL,NielsenFE,MetchiorT'JoenT'
Andereen PK. Does in-hospital ventricutarfibriltation affect prognosls after myocar-
dial infarctionl Eur Heon
l
1997 ;18:9 19
-924
lncidence of SCD after ventricutar fibrittation
9.
122a
Santoro
in acute
mwiven
prognosis ofearly primary ven_
sutts ofthe Gruppo ltatiano per
(GlSSl-2) database. AmlCordiol
10. Newby KH, Thompson T, Stebbins A, Topot EJ, Califf RM, Natale A. Sustained
ven_
tricurar arrhythmias in patients receiving thromborytic thempy: incidence and
outcomes. The GUSTO invesrigarors. Gl(ulotjon 199g:99:2567 _2573.
11.
p, poilackCVJr,
Z
q-
Harrington RA,
on compticating
*tu'";!l'^,
12.
tachy_
inter-
Sudden death in patients with myocardiar infarction and left ventricurar dysfunction,
heart failure, or both. N EnglJ rVed 2005;352:2581
-2588.
29. AdabagAS,TherneauTM,GershBj,WestonSA,RogerVL.Suddendeathafrermyo_
cardial nfarctio n. JA lvlA 2008;3 0O :2022 - 2029.
30. Hohntoser SH, Kuck KH, Dorian p, Roberts RS, Hampton
JR, Hatala R, Fain E,
Gent M, Connotly SJ. Prophytactic use ofan implantable cardioverter-defibrittator
i
aft er
acute myocardial infarction. N
I
E ngJ
Med ZOO4 ;!S
1
:24g1
-24g8,
31. Steinbeck G, Andresen D, Seidt K, BrachmannJ, Hoffmann E, Wo.jciechowski D,
Kornacewicz-.Jach Z, Sredniawa B, Lupkovics G, Hofgârtner F, others. Defibrillator
implantation early after myocardial infarction. N Engl./ /vled 2009:361:j427_1436.
32. Dorian P, Hohnloser SH, Thorpe KE, Robers RS, Kuck K-H, Gent M, Connolly SJ.
Mechanisms underlying the lack of erfect of imprantabte cardiovener-defibriilator
therapy on mortality in high-risk patients with recent myocardiat infarction: insights
from the defibrillation in acute myocardial infarction triat (DlNAMlr). arculàtion
2010;122:2645-2652.
33. Russo AM, Stainback RF, Bailey SR, Epstein AE, Heidenreich pA,
Jessup M, Kapa S,
Kremers MS, Lindsay BD, Stevenson LW. ACCF/HRS/AHA/ASE/HFSA/SCAI/
SCCT/SCMR 2013 appropriate use criteria for implantable cardioverterdefibritlators and cardiac resynchronization therapy: a report of the American
College of Cardiology Foundation appropriate use criteria task force, Heart
iogPhv
iety
34.
na-TomlA,
as-Cruz A,
farction
35.
16. Jouven
X
Desnos M, Guerot C, Ducimetière p. pædictingsudden
death in the popu_
lation: the paris prospective Study l. Circulonon f99;99:197g_j9g3.
17.
Beckman
e HL,
Dekker
Arnold
Zijlstra
primary
Weston SA, Meverden RA, Roger VL.
rdial infarction: a 2O-year community
36. Al-Khatib SM, Stebbins AL, Califi RM, Lee KL, Granger CB,
White HD,
Armstrong PW,TopolEJ, Ohman EM. Sustained ventricularairhythmias
and mortal_
ity among patients with acute myocardial infarction: results from
the
Heort I 2003;1 45;5 1 5-521.
An
KJ,
BJ, Catitr RM, Funk M, Kaufman ES, Krucoff MW Laks
MM, Macfarlane pW,
Sommargren C, SwirTn S, Van Hare GF, American Heart Association,
Councils on
Ca
Clinical
cular Disease in the
Yo
an
M
R,
s
for
ele
s
in
KS,
Nunes.Circulotrbn 20
3g.
eveît. circurotion
2006:114:1462-i467.
)0. Matsueys
i
and rethal v
2'1.
Councils on Cardio_
, and Cardiovascular
of Computerized Ete
an
Konelainen M-L, Linna E, Huikuri HV. Family history and the
risk of
sudden cardiac death as a manifestation of an acuce
coronary
,
g in hospital settings:
ciation
patients
19. Kaikkonen
GUSTollltrial.
37. Drew
others.
e causes; high
18.
H' clinicalimprica-
rions of an
vasosPastic angina
91 3.
Letter regarding article by piccini et ol, Sustained
cular fibriltarion complicating non-ST-segment_
es. Grculotion 201 3:l
tricular dysfunction, heart failure, or b oth. Circulotbn 2010;122:597 _602.
39. Sager PT, Perlmutter RA, Rosenfeld LE, Mcpherson CA, Wackers
FJ,
Batsford Wp
Electrophysiologic effects ofthrombolytic therapy in petients with
a transmural an_
terior myocardial infarction complicated by left ventricular aneue/sm formation
I An Coll Cordiot 1988:12:19 -24.
2f :e633.
22. Cambouj-BSimonT,MulakG,BatailleV,DanchinN,TheFrenchregistryofacuteST
elevation or non-ST_elevation myocardial infarction (FAST_Ml):
sltudy
baseline chamcteri stics. Arch Mol Cpeur Voiss 2007;l
00:524_534.
design and
24. Hinkle
LÊ Jr, Thater HT. Clinical classification of cardiac
deaths. Cjrculotjon 19g7;
65:457 -464.
25.
41. YanG-X.Phase2reentryæatriggertoinitiateventricularfibri{tationduringearty
acute myocardial ischemia. Circulotion 2004;110:1036_1041.
26.
77.
28
in
144-2151.
42.
, von der Thusen J
yocardiat infarction
diac
ffay
whi
card
etectroanatomic
A,
M,
rs,
43. Myerburg RJ,junttila
MJ.
culotion 2012:125:1043
a
Earty
char_
bdon
Sudden cardiac death caused by coronary heart disease. Gr_
-
1
052_